Red Post Collection: Gnar Discussion, Fearless on Sona Rework, Lyte on Prisoner's Island / Honor / AFK & Leavers, Recolor Survey and more!

Posted on at 1:10 AM by Moobeat
This morning's red post collect features Gypsylord and Meddler chatting about our upcoming champion Gnar, Fearless commenting on the upcoming Sona rework, Lyte discussing the concept of Prinsoner's Island and the honor system, and a peek at a recent Riot survey regarding the tentative idea of champion and skin recolors.
Continue reading for more information!


With Gnar's teaser going out earlier today, the forums have been suspiciously quiet with any additional details... although Gyspylord and Meddler did drop off a few comments.

Gypsylord, Gnar's designer, tweeted out that Gnar's ultimate is...
"His ultimate is incredibly powerful. Sometimes."

Meddler also popped into the forums to comment about Gnar's intended role, saying:
"We believe top lane's where Gnar's going to be played most due to the way his kit functions, though we've tested him in other roles too of course. As with any new release though that's just an estimate of course - internal testing's no substitute for the millions of games that'll get played once a champion's out that really dig into how they're best played."
As for how to pronounce Gnar, Meddler noted:
"Silent G."

When asked who did the gameplay work on Gnar, Meddler noted:
"Both Gypsylord and I have done gameplay design work on Gnar. Gypsylord's definitely done more of the work than I have though, particularly in the later parts of Gnar's development (most of my hands on contributions were in the earlier stages). CertainlyT's currently focused on another champion that's earlier in the production process.

Important to point out here though that gameplay design's just a piece of what goes into a champion however. A wide range of people are part of making a character like this (animators, visual effects artists, modellers, producers, QA, 2D artists, sound designers, writers, software engineers, playtesters from a wide range of disciplines etc) - gameplay design's just one piece of the puzzle."

As for details on Gnar's kit, Meddler teased:
"We'll get into kit discussion pretty soon, don't want to give any details away before the rest of the reveal stuff comes out though :)"

As for when they started working on Gnar, Meddler commented:
"Yeah, initial concept work on him started early last year. We figured a bunch of stuff out, then put him on ice for a while until we were ready to get him into the full production process."

Gypsylord also commented on Gnar and his relationship with Freljord
"Gnar does not currently reside in the Freljord (although that does not discount him possibly being "from" it). Honestly, I doubt he even has any understanding of what a "Freljord" is."

Fearless on Sona's rework and Feedback

With Sona's rework presumably headed to live patch 4.13, Fearless returned to the PBE boards to  clear the air on his absence from posting in the Sona rework feeback threads and offer context on the most recent tuning to her reworked kit:
"I wanted to post here to clear up the reasons we've been lowering Sona's values, even while the community here is asking for buffs. Most of the balance changes are reflections of internal playtesting, as we can balance those scenarios pretty well (accounting for people playing their main roles and champs, having less of a learning curve due to their closeness to the changes, etc.) so they give us a reasonably strong environment for understanding a champ's balance. 
PBE doesn't have all of this going for it. Matchmaking is much less precise on the PBE, and some players are just hopping into custom games.

Based off watching the internal playtests, Sona was strong, and not just a little strong. Sona really hadn't lost a lane before these changes.

As a side note, for the people claiming that negative posts don't get read, that's not true. I've read every Sona post I can find on here, on community sites, reddit, everything. While also trying to get Sona as streamlined and bug free as I can, time to reply is pretty lown When I see 3 or 4 people spamming out really angry posts, I do take the feedback and try to understand the source of frustration.

I haven't responded in this case because previous attempts to have discussions haven't proven productive. I don't expect to change the minds of people who've had poor experiences here based on talking about what happened in testing they can't see and don't believe in. Being called a monkey doesn't really make me believe that folks are interested in a meaningful discussion. If I have to choose between making an attempt to appease people who haven't displayed a tendency to actually interact with debate, or fixing up Sona for her release, I'm going to fix up Sona so that she's a better addition to the entire game, for as many players as I possibly can."
Fearless continued, again replying to criticism that he could have done more to communicate with the PBE community and reiterating the goal of the rework:
"I think it's a very fair criticism that I didn't speak to you all as much as you'd hoped for. Saying that only 3 comments on one thread doesn't suffice? That get's a bit tricky. What is the correct ratio of posts? Do I only count posts that have actually tried to open or inform a conversation, or should I count troll posts and people raging without having even played the rework? At what point is my time posting and searching for ambient Sona threads better spent talking with our QA, live balance team, and experimenting with changes, tweaks and bug fixing? It's very possible that I didn't get that formula even close to right. It's something for me to improve on.

That being said, when I look at posts that have lots of incorrect information, I started the time on PBE correcting them to make sure truthful information would be visible. Seeing those posts then get downvoted tells me that this community would rather punish someone they don't agree with rather than provide its members with the most accurate info. That's really depressing for me, both as a community member and a designer. It also doesn't help when words get put into my mouth. For E, example, I don't think the AP ratio "balances the ability." I think a completely new ability that gives Sona, as well as her allies a very large block of MS makes sense on a kit that focuses on positioning. Sona can now position for huge ults without Flash due to E, as well as use her other basic abilities with quicker repositioning to save her team and power them up.

When we change a champion, some of the balance will be comparative, but plenty of it can't be. Sona is Sona, and her balance as a whole is very different than any other champion. Sona has to be balanced of the collection as her base stats, abilities, growth, item paths, etc. I know it can be frustrating to see a value go down, and it can be easy to dwell on that change in isolation. When we're adjusting Sona, we're looking at the whole, and when she was showing up really strong, we tried to keep her power in the places that gave Sona her own set of strengths and weaknesses. That means she has to be worse at some things than other champions, but she gets to have strengths that are actually her own."
He continued, responding to another summoner who felt as if he did not address the PBE communities feedback:
"It's not that I don't care. Some of the suggestions don't fit the vision for the champion, some are just straight up OP, and some were tested and didn't work as well as the current option. It's been something I need to do better with to make sure that you and the rest of the community at least hears more in these situations. I'll admit to getting overwhelmed with the actual implementation and analysis for Sona, and not spending enough time addressing the feedback."
As for why Sona's aura where not increased above 350 range, he noted:
"The reason that we haven't increased the effective range of the auras, either through a straight range increase or through something like this, is that it is intended that Sona continuously has to make a choice between playing safe and playing to maximize the effect of every cast. This is also why Sona has gotten even better AP ratios. Sona players can build tanky/support tank and increase their safety, or they can opt into higher risk reward with AP builds. Larger effective aura sizes not only increase Sona's effective support output, but decrease interaction with allies and leave opponents no real chance to respond if she's strong, and no real room to buff her if she's weak."

Vi in the top lane

In a thread about Vi and why she isn't more prevalent in the top lane, Gypsylord, her designer, swung by the comment:
"So as anyone who has watched the Vi champ spotlight may know, Vi was originally intended to be a top laner.

She's got a lot of cool gameplay in top lane with the Q charge up and E cleave but, honestly, she's just too fair up there for too long compared to our current crop of top laners. In the early game she has meaningful mana costs, telegraphed attacks, and massive cooldown windows for her passive and all of her spells (P - 18, Q - 18, E - 14 per charge, R- 150). I actually think top lane Vi could be a pretty awesome experience for both sides of the lane if we did stuff like lower her CD's or buff E's range and mana cost but then jungle Vi would go to crazy town. Could nerf jungle Vi to compensate but then are we screwing over the many Vi mains who prefer her out-of-lane experience? It's a tough call. 
All this said, SmashGizmo got diamond last season almost exclusively with top lane Vi. He'd typically play conservative until 6 focusing only on CS and rush an Iceborn Gauntlet for the extended dueling power and mana to offset her high costs. Turns out if a Vi manages to stay even until 6 there is not a single character up there (not even Irelia or Jax) who can beat her in a straight up 1v1. Her burst and sustained damage output is insane when she gets the chance to use it and her late game with top lane items is monstrous for the enemy teams' squishies (or at least it was in season 3, a lot has changed since then)."

Lyte on "Prisoner's Island", Honor System, AFK/Leaver systems

Lyte returned to the forums to discuss the concept of "Prisoner's Island", the future updates to the Honor system, and the upcoming plans to address AFK/Leavers.

He started off with Prisoner's Island, a concept where toxic players are put in to their own separate queue and forced to play together.
"Hey Arkon Tri, 
The reason I haven't dug into this is because we've discussed why these designs aren't suitable for League (or many other games) in the past. Some other players have mentioned this, but you're basically suggesting a concept called "Prisoner's Island," where you pair players with similar reputations together.

There's a lot of nuances to low priority queue or Prisoner's Island systems; for example, you can use low priority queue systems for behaviors like Leaving/AFK. These behaviors are binary--players know they either left a game or not, and there's no subjectivity to the behavior. However, when you're dealing with behaviors like verbal abuse, intentional feeding or negative attitude, it's far more subjective. Some players will get placed into low priority queue systems for these behaviors and vehemently disagree that they belong there--you'll basically get a lot of, "What the? All these people in low priority queue are toxic, why am I here?
In these cases, a lot of players (mainly those without purchased content) will simply make new accounts to re-enter the ecosystem with a neutral reputation rating. If these players do that, you've basically banned a player but used a complicated, time-consuming system to do so. 
Secondly, a common issue with these designs is how you score new players. Do you score them in neutral priority, and force them to have worse matches than the "positive" players? This isn't ideal, because generally you want your new players to have the best game experience. You obviously can't start new players with 0 or negative reputation, as that'd be a negative experience. If you start new players with a positive reputation, then it gives toxic smurfs an easy way to ruin games at low levels while also further encouraging them to just start new accounts if they ever end up with negative reputation. 
We've talked about a lot more reasons in the past, but Prisoner's Island has basically been a concept for over 5 years and we know that it only works well with binary behaviors that are not subjective."
He continued:
"Ironically, what we find is that when you actually put players together in Prisoner's Island and all 10 players are actually consistently toxic players for subjective behaviors like verbal abuse... it's nearly impossible to change. 
Again, Prisoner's Island type systems work OK for behaviors like Leaving/AFKs which are binary, but are very, very counter-productive for subjective behaviors."

When asked about real rewards for positive players, Lyte mentioned the honor system and some upcoming plans:
"We have to go back and finish the Honor Initiative, and that'll explain a lot of the reasons why Honor was setup the way it is but seems incomplete. 
For example, right now there's an issue with the Honor Ribbons where it's simply rewarded to the top 20% of positive players on every server. Most of these players tend to play modes like ARAM and Co-op vs Bots because these modes simply have really high Honor give rates; so, you won't see ribbons as much in Normals or Ranked. One of the fixes is making sure that we reward ribbons per queue, and not just per server. 
Secondly, we have some rewards for the Honor Initiative that are more consistent. Right now, Ribbons are the top 20%, but there's nothing for the everyday player who wants to be recognized for being positive while working their way up."

As for ways to address Leavers/AFKs, Lyte noted:
"We've been working on a new system that addresses Leavers/AFKs more aggressively, and will talk about it in a post in the future. Without revealing too much, it's a different take on low priority designs and will update LeaverBuster's penalty and tuning systems."
He continued:
"Wrong, it has entirely new penalties and doesn't use bans."

In response to a player noting that anonymity is the source of players being toxic, he noted:
You will NEVER get rid of toxic players until you understand that anonymity is what causes it and that the only way you can reduce or get rid of it is to make it worth more to be nice then to be toxic. This system does just that. Your system does nothing but cause toxicity. Ever since it started the people you have banned have just made new accounts and are now raging on new player. Is that what your grand idea was? To have raging smurfs attacking new player?
Here's where you disagree with a lot of recent research by multiple university labs. A lot of recent research has shown that online toxicity is not about anonymity. If you look at news, social media and other games that force you to reveal your real ID and often personal profile to post comments, there's still an insane amount of toxicity. What the research has shown is that the problem is a lack of consequences for actions, and this has very little to do with anonymity. 
Secondly, the current systems don't create more toxicity. Many players that are banned for racism, sexism or homophobia quit League of Legends and never play again, and that is exactly what we wanted. They'll rage on the forums and complain, but that's expected when you tell a group of players they don't belong in this community. For the small number of players that do make new accounts, they'll keep getting banned until they quit as well. 
Third, there's actually a system in place at low levels already that separates toxic smurfs away from other players and then bans them. We knew some of the toxic players would make new accounts when they got banned, and we did want to protect new players. However, we see that for many toxic players a few bans does get them to quit permanently. Prisoner's Island systems don't solve this issue by the way... many players just create new accounts when they enter Prisoner's Island anyways. 
I agree that for most players, making it 'worth it' to be nice is more effective, and we've been working on efforts to reward those players and give them more reasons to reflect on why being positive = awesome. However, we're always going to need pretty aggressive systems for the small % of players that refuse to change. For these players, we just want to ban them repeatedly until they take their time elsewhere."

Riot Survey regarding Recolors

Over on reddit, a summoner shared a survey they had received from Riot regarding the idea of recolors for champion anad skin.

Riot BuboBubo stepped in to this thread to share context on the surveys and set some expectations for the highly tentative survey:
"Really happy to see the excitement here and wanted to provide a little bit more background. 
This survey is part of our efforts to understand what the most important underlying elements of “Recolors” are and how important they are relative to each other. Everything is super exploratory at this point (ie: far too early to commit to recolors at all, a specific feature set, price or release date). But this feedback will directly shape what gets built because ultimately we want to delight you and what better way to maximize the chances of achieving that than getting your input? 
On another note, surveys are a regular and central part of our quest to understand player wants, needs and behaviors. Importantly, they allow us to get representative responses from the various types of players. While forums and Reddit give us a lot of feedback and we are constantly perusing it, relying on them exclusively would lead to us missing the perspectives of players in other regions and those that don’t participate in these venues. This is also a key reason for why surveys are sent to individual players rather than posted for public comment."
When asked about these things popping up on Reddit and running wild , BuboBubo noted:
"Players who didn't get the survey enjoy the sneak peek, but it will inevitably set expectations for at least some players in ways that aren't ideal. Meeting, falling short of, or exceeding player expectations keeps us up at night and public posts definitely make things a bit more volatile. From the player insights perspective (my discipline at Riot), it doesn't help the validity of the surveys to have them posted up. 
But at the end of the day, its part of the risk and done without malicious intent so we'll roll with it."
In response to questions regarding a different survey than the recolor survey, BuboBubo noted:
"Its not uncommon for there to be multiple surveys being fielded at any given time. They're also used by a wide number of teams, including the ones that build team builder and skins. 
The 4 win IP bonus isn't the only reason to validate your email - it also gives you a higher likelihood of getting a sneak peek at ideas that are being considered."

You can see the questions asked in the survey below, thanks to a fan submission. Please, for the love of Teemo, heed my warning that this is highly tentative and NOT A CONFIRMED FEATURE OR COMMUNICATION REGARDING AN UPCOMING FEATURE.


No comments

Post a Comment